Does Adding a Flywheel Improve Edirtbike Performance

[DingusMcGee]

For starters an ebike motor essentially already has a flywheel. It is the heavy rotor of the BLDC motor. The simplest way to add more flywheel effect is to get more rotating mass on any part of the motor drive train that turn any time the motor rotor turns instead of making a bigger flywheel/rotor to fit(?) inside the motor case. The part of that drive train easiest to add flywheel effect with the tools I have is to add weight/mass to the rotating DSBBH (double sprocket bottom bracket hub.

Sprockets come in type A sprocket and type B sprocket . A type A sprocket is thin sheet of metal with teeth. A type B sprocket is a type A sprocket with a hub fastened to it. In a fashion the DSBBH hub is the combination of a type A sprocket fastened to a type B sprocket. When the DSBBH is fabricated with the larger sprocket type B, a larger rotating mass is achieved — more flywheel effect than making the DSBBH with the smaller sprocket type B.

See pics of the DSBBHs fabrications used in some of the tests:











These sprockets were made so as to be able to switch bearings for either a square tape BB spindle or an Isis BB spindle.

The physical quantity for the measurement of flywheel effect is the second moment of inertia (I). The measurement accounts for the observation that more rotational inertia is achieved as a given rotating mass gets further from the rotation center and this effect is non linear.

I am running out of time but will post the conclusion next post.

 

[DingusMcGee]

Performance improvement:

Background - these DSBBH flywheels (usually more than one Kg heavier than the lightweight version) were used on two edirtbikes. The motors were a QS1000 and the LR Big Block. The LR Big Block replaced a Cyclone 6k motor that never had a flywheel sprocket. My other 2 such bikes have a QS2000 V1 and a QS3000 V1 without added flywheels.

Riding Style - Not trials riding but more like tying to do the loop without stopping. These trails on hillsides are not manicured Mt Bike Trails but unmaintained trails used by gas motorized dirtbikes. They have sidehill ridding with a lot of loose scree/cobbles and headwalls while going uphill — difficult.

Even though the QS1000 and LR motors are small for riding the likes of such trails mentioned above, they do fine on smooth very steep trails. The flywheels on the right type of difficult terrain may have improved climbing performance by no more than 10%. They certainly help when encountering the first headwall on a steep scree trail. After the first headwall encounter the performance depends on how far it is to the next headwall or in other words from the previous. After any headwall encounter the motor needs a little time to gain rpm and put energy back into the flywheel. If speed is not regained the previous extra flywheel energy is not there. If the headwalls are too close for much rpm gain, you will need a bigger motor or more amperage output depending on batteries and controller capabilities.

These ebikes all have LiPo’s capable of some 24C discharge for the 20Ahr batts. The Fardriver controllers on them can output more than 150 amps as observered on some very smooth roadway accelerations — like going 75 mph with the QS3000. The QS 1000, LR Big Block and the Cyclone 6k just never exceeded about 80 amps when petering out on this steep difficult terrain. Before petering out, these motors go into the “Motor Fits” — jumping, sputtering, lerching and then stopping. Changing the startup PID settings on the controller does not help. The controller cannot get these small motors to respond sufficiently for the abruptly changing task at hand.

The solution seemed to get a bigger motor and so I replaced the QS1000 and LR Big Block with the new QS2000 V3. Now, no more problems with headwall after headwall on the trail with the bigger motor . I did not need to change any Fardriver settings. Bear in mine the QS2000V3 even with the gearbox weights less than the original QS2000V1 and has less max power.

For a summary the added flywheel performance is marginal. A bigger motor is more of an overall good investment for doing steep difficult trails than adding a flywheel to any smaller motor.

I weight about 150 lbs. If you clock in at 230 lbs, you likely will not get as much performance improvement.

 

[Chalo]

If you use a flywheel, it's folly to place it on any but the fastest rotating component. A slow flywheel is just ballast.

 

[DingusMcGee]

Hi Chalo,


Certainly KE = 0.5 x I x w**2 , so we have a non-linear energy gain with shaft rpm increases. But equally non-linear is effective distance of the rotating mass as the quantity I varies with r**2.

By your criteria I should have placed the flywheel directly on motor’s output shaft or created a high speed gearbox for more flywheel rpm than the motor shaft rpm?

Usually a quite small sprocket is used on the motor’s output shaft, so a type B sprocket would add little mass and the rotating effective radius is quite small with say a 12T sprocket. For this testing a fast turning sprocket with a big 2nd moment of inertia on the motor has little room and so a bigger second moment of inertia effect that could be achieved on the spindle could be incorporated without sacrificing precious space next to the motor shaft was employed.

To say doing other than what you stipulated is a folly is a folly in-itself as your declaration makes no quantitative assessment of what any trade-off design could achieve.

Furthermore if you had understood my assessment of the flywheel addition for driving up against multiple headwalls, you would be informed that the problem is getting energy back into the flywheel between headwall encounters. A small motor size cannot both grind uphill and restore enough flywheel energy for the next headwall. It is not that the bike needed a faster rotating flywheel like you suggest. Only with a bigger motor could the ebike/rider system successfully add enough energy for overcoming the pathway energy loss to do successive headwalls with trail roughness all along the pathway.